Overwatch and Asymmetric Level Design – What Makes the Maps Fun? – Extra Credits

Let’s talk about asymmetric level design, and the interesting case of Overwatch which for a lot of reasons has some of the most wildly asymmetric levels that I have seen in a multiplayer first-person shooter. [Intro music] Designing levels for competitive multiplayer games where the level design is asymmetric is a contentious thing. Even in games like MOBAs where the map is nearly symmetrical for both teams, minor minute differences, like which characters start near which monsters create a preference for players, and lead to players seeing one starting location as being more advantageous than the other. Meaning somebody’s always gonna be unhappy. And it’s a fine line for a designer to walk. Asymmetry provides variety and creates interesting tactical options for the player every time they play but it can also leave players just feeling frustrated and like the game is unfair every time they end up on the side they see as having a disadvantage. So first, let’s talk about standard asymmetric level design. Usually in asymmetric games, as the designer, you create all sorts of ways to make the action flow through the map. action flow through the map whether it’s Whether it’s a team deathmatch or capture the flag a team deathmatch or capture the flag you’ll notice that on any well designed map, you’ll notice that on any well designed map, the action doesn’t just occur at one point for the entire length of a match the action doesn’t just occur at one point for the entire length of a match. Instead, you’ll find yourself pushing through different choke points Instead, you’ll find yourself pushing through different choke points or dislodging your opponent from different sets of good cover. or dislodging your opponent from You’ll be taking different high ground areas or sniper spots as you go. you’ll be taking different high ground Sometimes these are subtle: sometimes these are subtle maybe you Maybe you simply have to push the enemy back enough simply have to push the enemy back to move forward to the next bit of cover. to move forward to the next bit of cover. Or sometimes these can be as drastic as pushing the enemy out of a set of buildings. Or sometimes these can be as drastic as pushing the enemy out of a set of buildings. But each time this happens, and players flow through the map, it does two important things. But each time this happens, and players flow through the map, it does two important things. First, the players get to think through new tactical problems. First, the players get to think through new tactical problems. Lines of sight are constantly changing, cover layouts change, Lines of sight are constantly changing, cover layouts change, height advantages change on different conflict points on the map. height advantages change on different This is interesting for the player! This is interesting for the player! It’s what makes the game feel vibrant and alive. it’s what makes the game feel vibrant Second, it constantly switches which team or players have an advantage. Second, it constantly switches which team or players have an advantage. Each new stopping point on the players’ circulation through the map Each new stopping point on the players’ circulation through the map should alter who gets an edge from the asymmetric terrain should alter who gets an edge from the asymmetric terrain which keeps the map feeling challenging but fair. which keeps the map feeling challenging but fair. There are a number of systems we use to achieve this: there are a number of systems we use to Baked into the level itself are conflict points. These are places everybody wants to be because they’re objectives or because they’re advantageous
these are places everybody wants to be like bunkers, hilltops, that sort of thing. Players will naturally gravitate toward those areas and the fight will naturally erupt around them as players try to take control As a battle erupts around one of these conflict points it will naturally drag other players in because that’s where the action is. Eventually the assaulting side will either push through eventually the assaulting side will or find the flank and push the possessing side back, or find the flank and push the possessing side back, bringing the flow of the game to whatever the next conflict point is as one side falls back and the other one surges forward. Or one of the sides may just give up the point as the fight becomes either no longer tenable or simply unwinnable. Either way, the conflict flows onto new areas. This is often augmented by the spawn system. There’s a lot of different philosophies around how to spawn the player into first person shooters. You could spawn them at a fixed location for the whole match. You could spawn them at one of a number of fixed spawn points. You could spawn them near a teammate. You could spawn them away from any action that’s going on. And there’s reasoning behind all of these options. In fact, just talking about what the different spawning mechanics do might make for an interesting episode one day. But for right now, the important thing is that, excluding the largely outmoded idea of spawning the player at a single fixed spawning location for an entire match, all of these options help the flow of play and encourages players to move to new conflict points. If the game has power ups or weapon pickups, these too can be used as conflict points and a way to drive the players from point to point so that the map feels varied but fair. But however you do it, your built-in map flow also has to depend on two other crucial questions. As a designer building an asymmetric map, you have to ask yourself: Are the players’ goals asymmetric and are the players avatars’ asymmetric? Because both of those have a huge effect on map design. And here is where we get to Overwatch. Because in Overwatch, the answer to both of those questions is yes. (at least most of the time) In Overwatch, there are basically two types of maps: In one type, one team is assaulting, trying to take ground and the other one is defending, trying to hold ground. In the other type, both teams are fighting to control a central objective. Interestingly, the second game type usually gives us maps that barely feel asymmetric at all as the design involves only one central conflict point with a group of minor conflict points radiating out from it Which often represent different paths to the central conflict point. Note that this is the only map type where they use a single fully fixed spawn location implying that the designers felt that asymmetry had far less impact on this map type and that they didn’t have to put as much emphasis on see-sawing between which team gets the most advantage out of the maps’ limited asymmetry. So let’s look at the attackers vs defenders maps, because those are wildly asymmetric. Intentionally so! The flow of each match through these maps is pretty evident. The designers channel the players’ movements much more clearly than your standard team deathmatch shooter. And because the designers know the map flow, this means that they can develop a type that creates a specific tension and interest curve for their online FPS. At least, theoretically The first thing to notice on all of these maps is that the attackers have the initial advantage. The first point the defenders are tasked with guarding usually gives the attacker lots of alternate paths to flank from and a modicum of cover. And the choke points are either relatively easy to break or can be avoided by smart flanking play On top of that, any defender who gets taken out has to hoof it all the way from their initial spawn point at the end of the map back to the area their team is trying to defend while the attackers respawn much closer to the initial conflict. Even if the defenders and the attackers suffer equal losses the respawn distance advantage will push the attack forward. But as points fall or carts are pushed, the situation flips. The attackers’ final goal usually sits at the end of one narrow corridor without any alternate routes to flank from and little cover to hide behind. And now the defenders have the respawn distance advantage as their spawn point is often right behind the goal Even trades now work in the defenders’ favor as their fallen teammates return to the fight in mere seconds. When this all works, it creates wonderfully nail-biting situations where the games come down to the last second. It makes both the attack and the defense feel like desperate hard-fought struggles while at the same time giving each match a sense of motion and progress. First the outer defenses collapse, and then the inner defenses fall until the defenders are left performing a desperate last stand on a patch of ground designed for desperate last stands. But! When you have this massive asymmetry that channels players down a specific path rather than allowing them to flow through the level naturally, you create a much more difficult balance problem. Because normally, when a particular conflict point is a bit too unbalanced in favor of one side or the other, players will naturally move their focus elsewhere and conflict will start again at some other point on the map. But that’s not possible in this case. Which means that the balance of Overwatch’s attacker-defender maps needs be tuned almost perfectly. And some of them are! Some of these levels are masterpieces of multiplayer design but there are others where that balance is just slightly off. Where the first point typically falls quickly and then the match turns into a seven-minute grind fest where the attackers throw themselves at an overly secure final defense point. Because these maps are designed to flow down one specific path, there’s no self-correcting mechanism in place to resolve the imbalance. Hopefully in time, the Overwatch team will be able to sand away those few remaining rough edges until all of the game’s levels live up to that perfectly crafted interest curve. It’s a daring and ambitious goal, one that has rarely, if ever, been pulled off. But they’ve got a lot of sharp people in that studio, so we’ll see what happens. But Overwatch’s asymmetry isn’t limited to level design. Next week, we’ll talk about the game’s most prominent asymmetric feature: Its characters. See ya then! (boop!) [Outro music]

About the author


  1. 'The first point the defenders defend from usually have multiple paths to go from'…

    Except Hanamura XD

    Seriosly that gate can be a pain to go through 🙁

  2. Did you just show tf2 while saying team deathmatch? IT DOES NOT HAVE TEAM DEATHMATCH

  3. alot of this can pertain to Battlefield 4 for example. play battlefield 4 for one match and look up. guarantee youll see people on a high rooftop and youll see Team A trying to keep team B from booting them off the roof and take it

  4. it's funny how when you said team deathmatch you put up a picture of tf2, even though tf2 is basically the opposite of that

  5. I personally believe that CS:GO has more asymmetric maps, although I can see that overwatch also has very good map design. Also, I like how every hero sort of has their own map.

  6. I like how this basically applies to tf2, but nope

    Overwatch is the thing now these days and made "Superior" Maps to the ones that have been working well for tf2 for the past 10 years

    And really? Team Death Match? TF2 has more modes than Overwatch, both Symmetic and Asymmetric designs

  7. I'm interested in a video about the different philosophies for spawning. The video says spawning at a fixed location is obsolete, but that is what Splatoon does, and it seems to be a successful game considering the relative failure of the Wii U

  8. I don't give a fuck. If I'm in Dorado and I don't spawn in the place with the bells which means I can't jam out, I'm pissed.

  9. Or just snowball the 2nd point so deaths are staggered, and get your whole team on the point with 3 or less enemies, and get a tic or two on the point of not taks it with 6 minutes for round 3 or 4 depending on who attacked first.

  10. Well, at least that's what they planned to have in OW.. When in reality typical OW game goes like this – attackers usually can't push through the very first choke point, cuz there is not enough (or NO AT ALL) flanking routes. Same objective has to be achieved through the same ways over and over again. Those maps with those gamemodes are just TOO MUCH linear, they become boring very fast. And if you get killed – you just respawn very far away from the combat and have to walk slowly back there again -> no action at all. Classic DM and TDM on the other hand don't have these issues. And thats cuz there is simply no fixed objective on maps, and maps themselves are not linear (a good DM map in fact has more than 2 ways to flank everywhere). In DM and TDM modes players create assymetry THEMSELVES, because they truely ARE the very objectives of the gamemode, and they constantly moving battleing each other. That means the map tactics always changes in real time, you have to always adapt. IMO, DM and TDM are WAY more interesting in that regard. So glad OW finaly has them (not to the extent i was expeting it to be though, but still)

  11. i think the map design in overwatch is a clusterfuck as well as the character design. it can be a fun game, but the bullshit just gets too stanky sometimes. the maps are cluttered to shit with random rooms and hallways and areas that just have no reason to exist and usually don't at all compliment the flow of the game or player movement. they're built solely on visual design, and the visual design is a clusterfuck too (can't see shit)

    i'm thinking of tf2's completely symmetrical 5CP map design and how it so well lays out three routes to the mid point, each with their own advantages and disadvantages. they're so carefully made and the flow of the game and player movement really banks off of that. the flow of 6v6 5CP tf2 is like a good chess game, and the flow of a competitive overwatch game is like monopoly on meth

  12. Tbh part of the reason why Overwatch is so boring after a while is that the final point is always too easy to defend. It narrows down the attackers too much and it limits what you can do. It doesn't become unwinnable but it gets stale because you only have one or two points of attack and so the defenders just pick tanks and healers and it becomes so difficult to win. The exception to this imo is Hanamora since the end point has the main entrance, the left entrance and the raised up areas to attack from giving lots of room for the attackers but I think that's because the first point has an easy to defend gate that once lost pretty much means the defenders should give up the point.

  13. Overwatch fails, not because of the levels, but because of the hero design. There is too much dependence on having a balanced team, that it becomes crippling when your teammates just want to play the popular heroes. And those heroes are? Well, they like to play the ones they lack the skill to use effectively, such as Genji and Widow, or they play the massively low skill ones, such as Junkrat and Mei. You have to beg the casuals to play tank or healer though. And if you play one of those yourself? Yeah, get ready for your shitty dps players to not kill anything.

    End result, is a game that tried to meld the brilliant playstyle of TF2, with Moba elements, and it sucks ass in every way (other than graphics and production quality). As much as I hate Paladins for taking the 'pander to the chinese P2W' approach, it is still better designed than Overwatch is.

    But hey, who cares, OWL is here now!

  14. The thing about everyone being salty about Team Fdeathmach 2 fuck up and how salty everyone is. Tf2 has a very high skill celing and a very vocal and tight nit fanbase, and those who play tf2 are very invested in it

  15. Is fun, it’s colorful, well designed and balanced and there are small Easter eggs and small details
    Except for Anubis
    It’s hell being on attack

  16. A good example for this is Splatoon’s Rainmaker and Tower Control. The Rainmaker and Tower Control start in the middle of the map, and when someone claims it, you can see that person on your map. Naturally, you will defend them or attack them. There’s always something going on around the tower or Rainmaker, which draws in people. You also have to claim these to make them move and get more points to win before the time runs out. The other team has to get to the point to obtain it, and the team that has it must defend it by keeping in groups.

  17. In War Thunder the terrain on maps us mostly natural landscape and even then they seem asymmetric. I definitely have favorite spawn points and sides on each map. It also has insane variation in tanks players play as in AB. The game can also be chaotic if when enemies flank into your teams area unseen.

  18. Honestly thought the TDM with the tf2 pic was a jab at matpat's deadlock video. don't know anymore after reading the comments.

  19. If you want to constantly use different characters, then u will want to make all your characters look cool. Over watch Cosmetics in a nutshell

  20. Honestly, most of OWs maps are ass. Better examples of good asymmetric design would've been Battlefield's Rush, Halo Reach's Invasion and of course TF2.

  21. 5:40

    hey one of the blue skulls has a lighter eye than the other ones, must have been a mista-

    wait a second…

    please don't tell me that's an undertale reference

Leave a Reply

Your email address will not be published. Required fields are marked *